@qa-gentic/agents 1.1.2

package/src/qa_stlc_agents/agent_helix_writer/tools/helix_write.py

@@ -0,0 +1,622 @@
+"""
+helix_write.py — File-system write tool for the Helix QA framework.
+
+Public API:
+  inspect_helix_project(helix_root)              -> dict
+  write_files_to_helix(helix_root, files, mode)  -> dict
+  read_helix_file(helix_root, relative_path)     -> dict
+  list_helix_tree(helix_root)                    -> dict
+
+  Framework existence detection
+  inspect_helix_project() returns framework_state: "absent"|"partial"|"present"
+  and a recommendation of "scaffold_and_tests" or "tests_only".
+  write_files_to_helix() accepts that recommendation as its `mode` argument.
+
+Infrastructure file protection
+  In mode="tests_only" all six utils/locators/*.ts files are always skipped.
+  In mode="scaffold_and_tests" they are written only if absent on disk,
+  unless force_scaffold=True is passed.
+
+Within-file deduplication
+  For locators.ts  : new const-object entries are merged; duplicate keys skipped.
+  For *.steps.ts   : new step blocks are appended; duplicate regex patterns skipped.
+  For *.page.ts    : new async methods are appended; duplicate method names skipped.
+
+Interface adapter
+  The generator emits repo.updateHealed / repo.incrementSuccess / repo.getBBox etc.
+  The existing Helix LocatorRepository only has setHealed / getBestSelector / getHealed.
+  _adapt_to_helix_interface() rewrites generated content before writing so it
+  compiles cleanly against the Helix interface without manual edits.
+"""
+from __future__ import annotations
+
+import re
+from pathlib import Path
+from typing import Any
+
+# ── Infrastructure file names ──────────────────────────────────────────────
+
+_INFRA_FILES = {
+    "LocatorHealer.ts",
+    "LocatorRepository.ts",
+    "TimingHealer.ts",
+    "VisualIntentChecker.ts",
+    "DevToolsHealer.ts",
+    "HealingDashboard.ts",
+}
+_REQUIRED_INFRA = {"LocatorHealer.ts", "LocatorRepository.ts"}
+_OPTIONAL_INFRA = _INFRA_FILES - _REQUIRED_INFRA
+
+_INFRA_RE    = re.compile(r"(LocatorHealer|LocatorRepository|TimingHealer|VisualIntentChecker|DevToolsHealer|HealingDashboard)\.ts$")
+_LOCATOR_RE  = re.compile(r"locators\.ts$",  re.IGNORECASE)
+_PAGE_RE     = re.compile(r"page\.ts$",      re.IGNORECASE)
+_STEPS_RE    = re.compile(r"steps\.ts$",     re.IGNORECASE)
+_FEATURE_RE  = re.compile(r"\.feature$",     re.IGNORECASE)
+_CUCUMBER_RE = re.compile(r"cucumber",       re.IGNORECASE)
+
+
+# ── Interface adapter ─────────────────────────────────────────────────────
+
+def _adapt_to_helix_interface(content: str) -> str:
+    """
+    Rewrite generated TypeScript so it compiles against the Helix-QA
+    LocatorRepository / LocatorHealer interface.
+
+    Generator emits        → Helix expects
+    ─────────────────────────────────────────────────────────────────
+    repo.updateHealed(k,s,…) → repo.setHealed(k, s)
+    repo.getBBox(key)         → null
+    repo.incrementSuccess(k)  → (removed)
+    repo.incrementFailure(k)  → (removed)
+    repo.queueSuggestion(…)   → (removed)
+    repo.updateBoundingBox(…) → (removed)
+    this.devtools.captureBoundingBox(…) → (removed)
+    fixture().logger          → this.logger
+    fixture().locatorRepository → this.repo
+    fixture().page            → this.page
+    import { Logger } from "winston" → import { HealerLogger }
+    import EnvironmentManager → import { environment } from @config/environment
+    new EnvironmentManager()  → environment
+    this.env.getBaseUrl()     → environment.getConfig().baseUrl
+    this.env.getPath('x')     → "x"
+    """
+    # repo.updateHealed(key, sel, ...) → repo.setHealed(key, sel)
+    content = re.sub(
+        r"this\.repo\.updateHealed\(\s*([^,)]+),\s*([^,)]+)(?:,[^)]+)?\)",
+        r"this.repo.setHealed(\1, \2)",
+        content,
+    )
+    # repo.getBBox(key) → null
+    content = re.sub(r"this\.repo\.getBBox\([^)]*\)", "null", content)
+
+    # Remove single-statement method calls that have no Helix equivalent
+    for method in ("incrementSuccess", "incrementFailure", "queueSuggestion",
+                   "updateBoundingBox", "captureBoundingBox"):
+        content = re.sub(
+            rf"^\s*(?:this\.repo\.|(?:this\.\w+\.)?){method}\([^)]*\)(?:\.catch\([^)]*\))?;?\s*\n",
+            "",
+            content,
+            flags=re.MULTILINE,
+        )
+
+    # fixture() references
+    content = re.sub(r"fixture\(\)\.logger\b",            "this.logger", content)
+    content = re.sub(r"fixture\(\)\.locatorRepository\b", "this.repo",   content)
+    content = re.sub(r"fixture\(\)\.page\b",              "this.page",   content)
+
+    # Winston Logger → Helix HealerLogger
+    content = content.replace(
+        'import { Logger } from "winston";',
+        'import { HealerLogger } from "./LocatorHealer";',
+    )
+    content = re.sub(r"\bLogger\b(?!\s*=)", "HealerLogger", content)
+
+    # EnvironmentManager → Helix environment singleton
+    content = content.replace(
+        'import { EnvironmentManager } from "@helper/environment/environmentManager.util";',
+        'import { environment } from "@config/environment";',
+    )
+    content = re.sub(r"\s*this\.env\s*=\s*new EnvironmentManager\(\);?\s*\n", "\n", content)
+    content = content.replace("new EnvironmentManager()", "environment")
+    content = content.replace("this.env.getBaseUrl()",    "environment.getConfig().baseUrl")
+    content = re.sub(r"this\.env\.getPath\(['\"]([^'\"]+)['\"]\)", r'"\1"', content)
+
+    return content
+
+
+# ── Path resolution ───────────────────────────────────────────────────────
+
+def _resolve_destination(helix_root: Path, file_key: str) -> Path:
+    key = file_key.strip()
+
+    if _CUCUMBER_RE.search(key):
+        return helix_root / "src" / "config" / "cucumber.config.ts"
+    if _INFRA_RE.search(key):
+        return helix_root / "src" / "utils" / "locators" / Path(key).name
+    if _FEATURE_RE.search(key):
+        return helix_root / "src" / "test" / "features" / Path(key).name
+    if _STEPS_RE.search(key):
+        return helix_root / "src" / "test" / "steps" / Path(key).name
+    if _LOCATOR_RE.search(key):
+        parts = Path(key).parts
+        stem = next(
+            (p for p in parts if p not in ("src", "pages", "locators", "utils") and not p.endswith(".ts")),
+            "page",
+        )
+        return helix_root / "src" / "locators" / f"{stem}.locators.ts"
+    if _PAGE_RE.search(key):
+        return helix_root / "src" / "pages" / Path(key).name
+
+    rel = key.lstrip("/")
+    return helix_root / (rel if rel.startswith("src/") else f"src/{rel}")
+
+
+# ── Within-file merge helpers ─────────────────────────────────────────────
+
+def _merge_locators(existing: str, generated: str) -> tuple[str, list[str], list[str]]:
+    """Append locator entries whose keys are not already in existing."""
+    existing_keys = set(re.findall(r"^\s{2}(\w+)\s*:", existing, re.MULTILINE))
+
+    new_entries: list[tuple[str, str]] = []
+    current_key: str | None = None
+    current_lines: list[str] = []
+
+    for line in generated.splitlines():
+        key_match = re.match(r"^  (\w+)\s*:\s*\{", line)
+        if key_match:
+            if current_key:
+                new_entries.append((current_key, "\n".join(current_lines)))
+            current_key = key_match.group(1)
+            current_lines = [line]
+        elif current_key:
+            current_lines.append(line)
+            if re.match(r"^\s*\},?\s*$", line):
+                new_entries.append((current_key, "\n".join(current_lines)))
+                current_key = None
+                current_lines = []
+
+    added: list[str] = []
+    skipped: list[str] = []
+    append_lines: list[str] = []
+
+    for key, block in new_entries:
+        if key in existing_keys:
+            skipped.append(key)
+        else:
+            append_lines.append(block)
+            added.append(key)
+
+    if not append_lines:
+        return existing, added, skipped
+
+    insertion = "\n" + "\n".join(append_lines)
+    merged = re.sub(r"(\n\}\s*as\s+const\s*;)", insertion + r"\1", existing, count=1)
+    if merged == existing:
+        merged = existing.rstrip() + "\n" + insertion + "\n"
+    return merged, added, skipped
+
+
+def _merge_steps(existing: str, generated: str) -> tuple[str, list[str], list[str]]:
+    """Append step blocks whose regex pattern is not already in existing."""
+    existing_patterns = set(re.findall(r"/\^([^/]+)\$/", existing))
+
+    step_block_re = re.compile(r"^(Given|When|Then)\(", re.MULTILINE)
+    parts = step_block_re.split(generated)
+
+    blocks: list[tuple[str, str]] = []
+    i = 1
+    while i + 1 < len(parts):
+        keyword = parts[i]
+        body    = parts[i + 1]
+        blocks.append((keyword, keyword + "(" + body))
+        i += 2
+
+    added:      list[str] = []
+    skipped:    list[str] = []
+    new_blocks: list[str] = []
+
+    for _kw, block in blocks:
+        pat_match = re.search(r"/\^([^/]+)\$/", block)
+        pattern = pat_match.group(1) if pat_match else block[:40]
+        if pattern in existing_patterns:
+            skipped.append(pattern)
+        else:
+            new_blocks.append(block)
+            added.append(pattern)
+
+    merged = existing.rstrip() + ("\n\n" + "\n".join(new_blocks) if new_blocks else "") + "\n"
+    return merged, added, skipped
+
+
+def _merge_page_methods(existing: str, generated: str) -> tuple[str, list[str], list[str]]:
+    """Append async methods whose names are not already in existing."""
+    existing_methods = set(re.findall(r"async\s+(\w+)\s*\(", existing))
+
+    method_re = re.compile(r"(  async\s+\w+\s*\([^)]*\)[^{]*\{)", re.MULTILINE)
+    raw_parts = method_re.split(generated)
+
+    method_blocks: list[tuple[str, str]] = []
+    i = 1
+    while i + 1 < len(raw_parts):
+        sig          = raw_parts[i]
+        body_and_rest = raw_parts[i + 1]
+        depth, end = 1, 0
+        for ch in body_and_rest:
+            if ch == "{":
+                depth += 1
+            elif ch == "}":
+                depth -= 1
+                if depth == 0:
+                    break
+            end += 1
+        method_blocks.append((sig, sig + body_and_rest[: end + 1]))
+        i += 2
+
+    added:       list[str] = []
+    skipped:     list[str] = []
+    new_methods: list[str] = []
+
+    for sig, block in method_blocks:
+        name_match = re.search(r"async\s+(\w+)\s*\(", sig)
+        name = name_match.group(1) if name_match else sig[:30]
+        if name in existing_methods:
+            skipped.append(name)
+        else:
+            new_methods.append(block)
+            added.append(name)
+
+    if not new_methods:
+        return existing, added, skipped
+
+    insertion = "\n" + "\n\n".join(new_methods) + "\n"
+    merged = re.sub(r"\n\}\s*\n?$", insertion + "\n}\n", existing, count=1)
+    if merged == existing:
+        merged = existing.rstrip() + insertion + "}\n"
+    return merged, added, skipped
+
+
+# ── Framework state inspection ────────────────────────────────────────────
+
+def inspect_helix_project(helix_root: str) -> dict[str, Any]:
+    """
+    Examine the Helix-QA project root and return a machine-readable verdict.
+
+    framework_state: "absent" | "partial" | "present"
+    recommendation:  "scaffold_and_tests" | "tests_only"
+
+    "absent"   helix_root missing, or src/ absent, or both required infra files absent.
+    "partial"  src/ exists but one or both required infra files missing.
+    "present"  Both LocatorHealer.ts and LocatorRepository.ts exist on disk.
+    """
+    root = Path(helix_root).expanduser().resolve()
+
+    if not root.exists():
+        return {
+            "framework_state": "absent",
+            "missing_infra": sorted(_INFRA_FILES),
+            "existing_infra": [],
+            "has_src": False,
+            "recommendation": "scaffold_and_tests",
+            "message": (
+                f"helix_root '{root}' does not exist. "
+                "Create the directory and use mode='scaffold_and_tests'."
+            ),
+        }
+
+    infra_dir = root / "src" / "utils" / "locators"
+    has_src   = (root / "src").exists()
+
+    existing_infra   = sorted(f for f in _INFRA_FILES     if (infra_dir / f).exists())
+    missing_infra    = sorted(f for f in _INFRA_FILES     if f not in existing_infra)
+    missing_required = sorted(f for f in _REQUIRED_INFRA  if f not in existing_infra)
+
+    if not has_src or missing_required:
+        state = "absent" if not has_src else "partial"
+        return {
+            "framework_state": state,
+            "missing_infra": missing_infra,
+            "existing_infra": existing_infra,
+            "has_src": has_src,
+            "recommendation": "scaffold_and_tests",
+            "message": (
+                f"Helix-QA framework is {state}. "
+                f"Missing required files: {missing_required}. "
+                "Run scaffold_locator_repository, then write_helix_files with mode='scaffold_and_tests'."
+            ),
+        }
+
+    return {
+        "framework_state": "present",
+        "missing_infra": missing_infra,
+        "existing_infra": existing_infra,
+        "has_src": True,
+        "recommendation": "tests_only",
+        "message": (
+            "Helix-QA framework is present. "
+            "Infrastructure files will not be touched. "
+            "Only test files will be written or merged."
+        ),
+    }
+
+
+# ── Main write function ────────────────────────────────────────────────────
+
+def write_files_to_helix(
+    helix_root:     str,
+    files:          dict[str, str],
+    mode:           str  = "scaffold_and_tests",
+    force_scaffold: bool = False,
+) -> dict[str, Any]:
+    """
+    Write generated files into the Helix-QA project with full deduplication.
+
+    mode="tests_only"          write locators, pages, steps, features — merge
+                               into existing files, skip infra files always.
+    mode="scaffold_and_tests"  also write infra files that do not yet exist.
+    force_scaffold=True        overwrite existing infra files (use deliberately).
+    """
+    root = Path(helix_root).expanduser().resolve()
+    if not root.exists():
+        return {
+            "success": False,
+            "error": f"helix_root does not exist: {root}",
+            "written": [], "skipped": [],
+        }
+
+    written:       list[dict] = []
+    skipped:       list[dict] = []
+    deduplication: dict[str, Any] = {}
+
+    for file_key, raw_content in files.items():
+        if not raw_content or not raw_content.strip():
+            skipped.append({"file_key": file_key, "reason": "empty content"})
+            continue
+
+        # Apply interface adapter to all TypeScript content
+        content = _adapt_to_helix_interface(raw_content)
+
+        dest     = _resolve_destination(root, file_key)
+        dest.parent.mkdir(parents=True, exist_ok=True)
+        dest_rel = str(dest.relative_to(root))
+
+        # ── Infrastructure files ───────────────────────────────────────────
+        if _INFRA_RE.search(file_key):
+            if mode == "tests_only":
+                skipped.append({
+                    "file_key": file_key, "dest": dest_rel,
+                    "reason": "infrastructure file — skipped in tests_only mode",
+                })
+                continue
+            if dest.exists() and not force_scaffold:
+                skipped.append({
+                    "file_key": file_key, "dest": dest_rel,
+                    "reason": "infrastructure file already exists (pass force_scaffold=True to overwrite)",
+                })
+                continue
+            try:
+                action = "overwritten" if dest.exists() else "created"
+                dest.write_text(content, encoding="utf-8")
+                written.append({"file_key": file_key, "dest": dest_rel,
+                                 "bytes": len(content.encode()), "action": action})
+            except OSError as exc:
+                skipped.append({"file_key": file_key, "dest": dest_rel, "reason": str(exc)})
+            continue
+
+        # ── Cucumber config: append profile, skip duplicate ────────────────
+        if _CUCUMBER_RE.search(file_key):
+            try:
+                if dest.exists():
+                    existing_text  = dest.read_text(encoding="utf-8")
+                    profile_match  = re.match(r"\s*(\w+)\s*:", content.strip())
+                    profile_name   = profile_match.group(1) if profile_match else None
+                    if profile_name and profile_name in existing_text:
+                        skipped.append({
+                            "file_key": file_key, "dest": dest_rel,
+                            "reason": f"profile '{profile_name}' already exists in cucumber.config.ts",
+                        })
+                        continue
+                    dest.write_text(
+                        existing_text.rstrip() + "\n\n// --- generated profile ---\n" + content,
+                        encoding="utf-8",
+                    )
+                    written.append({"file_key": file_key, "dest": dest_rel,
+                                    "bytes": len(content.encode()), "action": "appended"})
+                else:
+                    dest.write_text(content, encoding="utf-8")
+                    written.append({"file_key": file_key, "dest": dest_rel,
+                                    "bytes": len(content.encode()), "action": "created"})
+            except OSError as exc:
+                skipped.append({"file_key": file_key, "dest": dest_rel, "reason": str(exc)})
+            continue
+
+        # ── Feature files: overwrite (they are the Gherkin source of truth) ─
+        if _FEATURE_RE.search(file_key):
+            try:
+                action = "overwritten" if dest.exists() else "created"
+                dest.write_text(content, encoding="utf-8")
+                written.append({"file_key": file_key, "dest": dest_rel,
+                                 "bytes": len(content.encode()), "action": action})
+            except OSError as exc:
+                skipped.append({"file_key": file_key, "dest": dest_rel, "reason": str(exc)})
+            continue
+
+        # ── Merge-aware write for locators / page / steps ─────────────────
+        try:
+            if not dest.exists():
+                dest.write_text(content, encoding="utf-8")
+                written.append({"file_key": file_key, "dest": dest_rel,
+                                 "bytes": len(content.encode()), "action": "created"})
+                continue
+
+            existing_text = dest.read_text(encoding="utf-8")
+
+            if _LOCATOR_RE.search(file_key):
+                merged, added, dup = _merge_locators(existing_text, content)
+                deduplication[dest_rel] = {
+                    "type": "locators", "added_keys": added, "skipped_keys": dup,
+                }
+            elif _STEPS_RE.search(file_key):
+                merged, added, dup = _merge_steps(existing_text, content)
+                deduplication[dest_rel] = {
+                    "type": "steps", "added_patterns": added, "skipped_patterns": dup,
+                }
+            elif _PAGE_RE.search(file_key):
+                merged, added, dup = _merge_page_methods(existing_text, content)
+                deduplication[dest_rel] = {
+                    "type": "page", "added_methods": added, "skipped_methods": dup,
+                }
+            else:
+                merged = content
+                deduplication[dest_rel] = {"type": "unknown", "action": "overwritten"}
+
+            dest.write_text(merged, encoding="utf-8")
+            written.append({"file_key": file_key, "dest": dest_rel,
+                             "bytes": len(merged.encode()), "action": "merged"})
+
+        except OSError as exc:
+            skipped.append({"file_key": file_key, "dest": dest_rel, "reason": str(exc)})
+
+    return {
+        "success": len(written) > 0 or len(skipped) == 0,
+        "helix_root": str(root),
+        "written": written,
+        "skipped": skipped,
+        "deduplication": deduplication,
+        "summary": {
+            "requested": len(files),
+            "written":   len(written),
+            "skipped":   len(skipped),
+        },
+    }
+
+
+# ── Read / list helpers ───────────────────────────────────────────────────
+
+def read_helix_file(helix_root: str, relative_path: str) -> dict[str, Any]:
+    root   = Path(helix_root).expanduser().resolve()
+    target = (root / relative_path.lstrip("/")).resolve()
+    try:
+        target.relative_to(root)
+    except ValueError:
+        return {"success": False, "error": "Path escapes helix_root"}
+    if not target.exists():
+        return {"success": False, "exists": False, "path": relative_path}
+    try:
+        content = target.read_text(encoding="utf-8")
+        return {"success": True, "exists": True, "path": relative_path,
+                "content": content, "bytes": len(content.encode())}
+    except OSError as exc:
+        return {"success": False, "error": str(exc)}
+
+
+def update_helix_file(
+    helix_root:     str,
+    relative_path:  str,
+    content:        str,
+    force_overwrite: bool = False,
+) -> dict[str, Any]:
+    """
+    Write or merge a single file in the Helix-QA project.
+
+    Applies the same interface adapter and merge logic as write_files_to_helix
+    but targets one file directly, avoiding a full-dict write call.
+
+    force_overwrite=True  — replace the entire file without merging.
+    """
+    root   = Path(helix_root).expanduser().resolve()
+    target = (root / relative_path.lstrip("/")).resolve()
+    # Prevent path traversal
+    try:
+        target.relative_to(root)
+    except ValueError:
+        return {"success": False, "error": "relative_path escapes helix_root — path traversal rejected"}
+
+    if not content or not content.strip():
+        return {"success": False, "error": "content is empty — nothing to write"}
+
+    adapted = _adapt_to_helix_interface(content)
+    target.parent.mkdir(parents=True, exist_ok=True)
+    dest_rel = str(target.relative_to(root))
+
+    try:
+        if not target.exists() or force_overwrite:
+            action = "overwritten" if target.exists() else "created"
+            target.write_text(adapted, encoding="utf-8")
+            return {
+                "success": True,
+                "path": dest_rel,
+                "action": action,
+                "bytes": len(adapted.encode()),
+                "deduplication": None,
+            }
+
+        existing_text = target.read_text(encoding="utf-8")
+        file_key = relative_path
+
+        if _LOCATOR_RE.search(file_key):
+            merged, added, dup = _merge_locators(existing_text, adapted)
+            dedup = {"type": "locators", "added_keys": added, "skipped_keys": dup}
+        elif _STEPS_RE.search(file_key):
+            merged, added, dup = _merge_steps(existing_text, adapted)
+            dedup = {"type": "steps", "added_patterns": added, "skipped_patterns": dup}
+        elif _PAGE_RE.search(file_key):
+            merged, added, dup = _merge_page_methods(existing_text, adapted)
+            dedup = {"type": "page", "added_methods": added, "skipped_methods": dup}
+        elif _FEATURE_RE.search(file_key):
+            # Feature files are the Gherkin source of truth — always overwrite
+            target.write_text(adapted, encoding="utf-8")
+            return {
+                "success": True,
+                "path": dest_rel,
+                "action": "overwritten",
+                "bytes": len(adapted.encode()),
+                "deduplication": None,
+            }
+        else:
+            merged = adapted
+            dedup  = {"type": "unknown", "action": "overwritten"}
+
+        target.write_text(merged, encoding="utf-8")
+        return {
+            "success": True,
+            "path": dest_rel,
+            "action": "merged",
+            "bytes": len(merged.encode()),
+            "deduplication": dedup,
+        }
+
+    except OSError as exc:
+        return {"success": False, "error": str(exc), "path": dest_rel}
+
+
+def list_helix_tree(helix_root: str) -> dict[str, Any]:
+    root = Path(helix_root).expanduser().resolve()
+    if not root.exists():
+        return {"success": False, "error": f"helix_root does not exist: {root}"}
+
+    tree: dict[str, list[str]] = {
+        "features": [], "steps": [], "pages": [],
+        "locators": [], "utils_locators": [], "other": [],
+    }
+    for path in sorted(root.rglob("*")):
+        if not path.is_file():
+            continue
+        if path.suffix not in (".ts", ".feature", ".js"):
+            continue
+        if any(p in ("node_modules", "dist", "test-results", ".git") for p in path.parts):
+            continue
+        rel = str(path.relative_to(root))
+        if "test/features" in rel or rel.endswith(".feature"):
+            tree["features"].append(rel)
+        elif "test/steps" in rel and rel.endswith(".ts"):
+            tree["steps"].append(rel)
+        elif "src/pages" in rel and rel.endswith(".ts"):
+            tree["pages"].append(rel)
+        elif "src/locators" in rel and rel.endswith(".ts"):
+            tree["locators"].append(rel)
+        elif "utils/locators" in rel and rel.endswith(".ts"):
+            tree["utils_locators"].append(rel)
+        else:
+            tree["other"].append(rel)
+
+    return {"success": True, "helix_root": str(root), "tree": tree}